In [36]:
import pandas as pd
In [2]:
df_test=pd.read_csv("C:/Users/Anderson/Downloads/UNSW-NB15 - CSV Files/UNSW-NB15 - CSV Files/a part of training and testing set/UNSW_NB15_training-set.csv")
df_test=df_test.drop(['id'], axis=1)
df_train=pd.read_csv("C:/Users/Anderson/Downloads/UNSW-NB15 - CSV Files/UNSW-NB15 - CSV Files/a part of training and testing set/UNSW_NB15_testing-set.csv")
df_train=df_train.drop(['id'], axis=1)
df=pd.concat([df_train,df_test])
In [3]:
df.shape
df.head()
Out[3]:
In [4]:
import seaborn as s
import matplotlib.pyplot as plt
%matplotlib inline
plt.figure(figsize=(4,8))
ax=s.countplot(x='label',data=df_train )
for p in ax.patches:
ax.annotate('{:}'.format(p.get_height()), (p.get_x()+0.1, p.get_height()+50))
In [5]:
import seaborn as s
import matplotlib.pyplot as plt
%matplotlib inline
plt.figure(figsize=(4,8))
ax=s.countplot(x='label',data=df_test )
for p in ax.patches:
ax.annotate('{:}'.format(p.get_height()), (p.get_x()+0.1, p.get_height()+50))
In [6]:
import seaborn as s
import matplotlib.pyplot as plt
%matplotlib inline
plt.figure(figsize=(4,8))
ax=s.countplot(x='label',data=df )
for p in ax.patches:
ax.annotate('{:}'.format(p.get_height()), (p.get_x()+0.1, p.get_height()+50))
In [25]:
s.pairplot(df, hue=None, hue_order=None, palette=None, vars=None, x_vars=None, y_vars=None, kind='scatter', diag_kind='hist', markers=None, size=2.5, aspect=1, dropna=True, plot_kws=None, diag_kws=None, grid_kws=None)
Out[25]:
In [7]:
df=df.drop(['attack_cat'], axis=1)
In [8]:
from sklearn.preprocessing import LabelEncoder,OneHotEncoder
df_categorical_values_enc=df.select_dtypes(include=['object']).apply(LabelEncoder().fit_transform)
print(df_categorical_values_enc.head())
# proto
proto=sorted(df.proto.unique())
string1 = 'Proto_'
proto_2=[string1 + x for x in proto]
# service
service=sorted(df.service.unique())
string2 = 'service_'
service_2=[string2 + x for x in service]
# state
state=sorted(df.state.unique())
string3 = 'state_'
state_2=[string3 + x for x in state]
# put together
dumcols=[]
#dumcols=''.join(proto_2 + service_2 + state_2)
dumcols.append(proto_2)
dumcols.append(service_2)
dumcols.append(state_2)
print(dumcols)
def flatten(l):
try:
return flatten(l[0]) + (flatten(l[1:]) if len(l) > 1 else []) if type(l) is list else [l]
except IndexError:
return []
dumcols=flatten(dumcols)
print(dumcols)
len(dumcols)
Out[8]:
In [9]:
enc = OneHotEncoder()
df_categorical_values_encenc = enc.fit_transform(df_categorical_values_enc)
df_cat_data = pd.DataFrame(df_categorical_values_encenc.toarray(),columns=dumcols)
df_cat_data.head()
Out[9]:
In [10]:
newdf=df.join(df_cat_data)
newdf.drop('state', axis=1, inplace=True)
newdf.drop('proto', axis=1, inplace=True)
newdf.drop('service', axis=1, inplace=True)
print(newdf.shape)
In [11]:
newdf.head()
Out[11]:
In [12]:
colnames=newdf.columns
In [13]:
X=newdf.drop(['label'], axis=1)
y=newdf.label
In [14]:
colNames=colnames
In [15]:
from sklearn.feature_selection import RFE
from sklearn.ensemble import RandomForestClassifier
clf = RandomForestClassifier(n_jobs=2)
rfe = RFE(estimator=clf, n_features_to_select=25, step=1)
rfe.fit(X, y)
X_rfe=rfe.transform(X)
true=rfe.support_
rfecolindex=[i for i, x in enumerate(true) if x]
rfecolname=list(colNames[i] for i in rfecolindex)
print('Features selected for train:',rfecolname)
print()
In [16]:
from sklearn import preprocessing
scaler_train = preprocessing.StandardScaler().fit(X)
X=scaler_train.transform(X)
print(X.std(axis=0))
In [17]:
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)
In [18]:
from sklearn.decomposition import PCA
pca = PCA(n_components=2)
X_train = pca.fit_transform(X_train)
X_test = pca.transform(X_test)
explained_variance = pca.explained_variance_ratio_
In [19]:
from sklearn.linear_model import LogisticRegression
classifier = LogisticRegression(random_state = 0)
classifier.fit(X_train, y_train)
Out[19]:
In [20]:
y_pred = classifier.predict(X_test)
print (y_pred)
print(y_test)
In [21]:
from matplotlib.colors import ListedColormap
import numpy as np
X_set, y_set = X_train, y_train
X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01),
np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.01))
plt.contourf(X1, X2, classifier.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),
alpha = 0.75, cmap = ListedColormap(('red', 'green', 'blue')))
plt.xlim(X1.min(), X1.max())
plt.ylim(X2.min(), X2.max())
for i, j in enumerate(np.unique(y_set)):
plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],
c = ListedColormap(('red', 'green', 'blue'))(i), label = j)
plt.title('Logistic Regression (Training set)')
plt.xlabel('PC1')
plt.ylabel('PC2')
plt.legend()
plt.show()
In [22]:
from sklearn.ensemble import RandomForestClassifier
clf=RandomForestClassifier(n_estimators=10, criterion='gini', max_depth=None, min_samples_split=2,
min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_features='auto',
max_leaf_nodes=None, min_impurity_decrease=0.0, min_impurity_split=None,
bootstrap=True, oob_score=False, n_jobs=1, random_state=None, verbose=0,
warm_start=False, class_weight=None)
In [23]:
clf.fit(X_train,y_train)
Out[23]:
In [24]:
y_pred=clf.predict(X_test)
In [25]:
pd.crosstab(y_test, y_pred, rownames=['Actual attacks'], colnames=['Predicted attacks'])
Out[25]:
In [26]:
from sklearn.metrics import classification_report
print("Model evaluation\n"+classification_report(y_test,y_pred))
from sklearn.model_selection import cross_val_score
accuracy = cross_val_score(clf, X_test, y_test, cv=10, scoring='accuracy')
print(accuracy)
print("Accuracy: %0.9f (+/- %0.9f)" % (accuracy.mean(), accuracy.std() * 2))
In [27]:
from matplotlib.colors import ListedColormap
import numpy as np
X_set, y_set = X_train, y_train
X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01),
np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.01))
plt.contourf(X1, X2, clf.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),
alpha = 0.75, cmap = ListedColormap(('red', 'green', 'blue')))
plt.xlim(X1.min(), X1.max())
plt.ylim(X2.min(), X2.max())
for i, j in enumerate(np.unique(y_set)):
plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],
c = ListedColormap(('red', 'green', 'blue'))(i), label = j)
plt.title('random forest (Training set)')
plt.xlabel('PC1')
plt.ylabel('PC2')
plt.legend()
plt.show()
In [28]:
from sklearn import svm
clf = svm.SVC()
In [29]:
clf.fit(X_train,y_train)
y_pred=clf.predict(X_test)
In [30]:
pd.crosstab(y_test, y_pred, rownames=['Actual attacks'], colnames=['Predicted attacks'])
Out[30]:
In [ ]:
from matplotlib.colors import ListedColormap
import numpy as np
X_set, y_set = X_train, y_train
X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01),
np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.01))
plt.contourf(X1, X2, clf.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),
alpha = 0.75, cmap = ListedColormap(('red', 'green', 'blue')))
plt.xlim(X1.min(), X1.max())
plt.ylim(X2.min(), X2.max())
for i, j in enumerate(np.unique(y_set)):
plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],
c = ListedColormap(('red', 'green', 'blue'))(i), label = j)
plt.title('SVM (Training set)')
plt.xlabel('PC1')
plt.ylabel('PC2')
plt.legend()
plt.show()
In [32]:
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)
In [34]:
from sklearn.ensemble import RandomForestClassifier
clf=RandomForestClassifier(n_estimators=10, criterion='gini', max_depth=None, min_samples_split=2,
min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_features='auto',
max_leaf_nodes=None, min_impurity_decrease=0.0, min_impurity_split=None,
bootstrap=True, oob_score=False, n_jobs=1, random_state=None, verbose=0,
warm_start=False, class_weight=None)
clf.fit(X_train,y_train)
y_pred=clf.predict(X_test)
pd.crosstab(y_test, y_pred, rownames=['Actual attacks'], colnames=['Predicted attacks'])
Out[34]:
In [35]:
from sklearn.metrics import classification_report
print("Model evaluation\n"+classification_report(y_test,y_pred))
from sklearn.model_selection import cross_val_score
accuracy = cross_val_score(clf, X_test, y_test, cv=10, scoring='accuracy')
print(accuracy)
print("Accuracy: %0.9f (+/- %0.9f)" % (accuracy.mean(), accuracy.std() * 2))